Researcher Profile

Featured

Nicholas Carlini

Adversarial ML, security of deployed models

Researcher at Anthropic

One of the most useful people to study if you care about what deployed models get wrong under pressure, especially around extraction, adversarial behavior, and practical security failures.

Organizations

Anthropic

Topics

Post-Training & Alignment Evaluation & Benchmarks Security & Robustness

About This Page

This profile is meant to help you get oriented quickly: why this researcher matters, what to read first, and where to explore next.

Last reviewed

March 18, 2026

Best First Clicks

Towards Evaluating the Robustness of Neural Networkspaper Stealing Machine Learning Models via Prediction APIspaper Extracting Training Data from Large Language Modelspaper

Known For

The ideas, systems, and research directions that make this person worth knowing.

Adversarial ML and extraction risks

Security research on deployed models

Practical failure modes in modern ML systems

Adversarial ML, security of deployed models

Nicholas Carlini (site)

Security

Start Here

Canonical papers, project pages, or repositories that anchor this profile.

Towards Evaluating the Robustness of Neural Networkspaper Stealing Machine Learning Models via Prediction APIspaper Extracting Training Data from Large Language Modelspaper Nicholas Carlini (site)website

Supporting Sources

Additional links that help verify and flesh out this profile.

Nicholas Carlini career updatearticle

Related Researchers

People worth exploring next because they share topics, labs, or source material with this profile.

Shared topics

Francis Song

Red teaming with language models

1 source

Co-authored Red Teaming LMs with LMs: a concrete approach to stress-testing model behavior at scale.

Post-Training & Alignment Evaluation & Benchmarks

Start HereRed Teaming Language Models with Language Models

Shared topics

John Aslanides

Red teaming with language models

1 source

Co-authored Red Teaming LMs with LMs: a concrete approach to stress-testing model behavior at scale.

Post-Training & Alignment Evaluation & Benchmarks

Start HereRed Teaming Language Models with Language Models

Shared topics

Nat McAleese

Red teaming with language models

1 source

Co-authored Red Teaming LMs with LMs: a concrete approach to stress-testing model behavior at scale.

Post-Training & Alignment Evaluation & Benchmarks

Start HereRed Teaming Language Models with Language Models

Shared topics

Saffron Huang

Red teaming with language models

1 source

Co-authored Red Teaming LMs with LMs: a concrete approach to stress-testing model behavior at scale.

Post-Training & Alignment Evaluation & Benchmarks

Start HereRed Teaming Language Models with Language Models

Shared canonical source

Alina Oprea

Training-data extraction and privacy risks

1 source

Co-authored Extracting Training Data from Large Language Models: a core paper on memorization and extraction risk.

Security & Robustness

Start HereExtracting Training Data from Large Language Models

Shared canonical source

Eric Wallace

Training-data extraction and privacy risks

1 source

Co-authored Extracting Training Data from Large Language Models: a core paper on memorization and extraction risk.

Security & Robustness

Start HereExtracting Training Data from Large Language Models